Debonding of Dutch intensifying prefixoids: a multiple source account Kristel Van Goethem, Muriel Norde F.R.S.-FNRS & Université catholique de Louvain; Humboldt University, Berlin Objectives In this study, we will investigate how Dutch affixes and affixoids develop autonomous uses through debonding, i.e. "a composite change whereby a bound morpheme in a specific linguistic context becomes a free morpheme" (Norde 2009: 186). Debonding is one of the three subtypes of degrammaticalization distinguished by Norde (2009) and is therefore characterized as a gradual and construction-internal process. More specifically, debonding typically involves severance (i.e. decrease in bondedness), flexibilization (i.e. increase in syntactic freedom), scope expansion and possibly recategorialization and resemanticization. The focus of our study will be on Dutch intensifying prefixoids (e.g. een topjaar 'a top year', een reuzehonger 'a giant appetite', reuzeaardig ‘lit. giant nice; very nice’, keigezellig ‘lit. boulder cosy; very cosy’) which develop adjectival and/or adverbial uses by debonding. Hypothesis Debonding applies to clitics, inflectional and derivational affixes (Norde 2009: 186-227). As has been shown in many studies (among others the ones collected in Dressler et al. (Eds) 2005 and Scalise & Vogel (Eds) 2010), there is however no strict demarcation between compounding and derivation. Dutch explicitly manifests this gradience by a productive system of so-called affixoids: i.e. "morphemes which look like parts of compounds, and do occur as lexemes, but have a specific and more restricted meaning when used as part of a compound" (Booij 2009: 208). Affixoids are not true affixes, because they still have lexical counterparts (cf. reuzeblij 'lit. giant-happy; very happy' vs de reus Goliath 'Goliath the giant'), but they undergo semantic specialization when used as bound forms and then mostly develop a more abstract meaning, for instance intensification, in which case they are comparable to intensifying affixes (cf. reuzeblij / superblij 'very happy'). Since affixoids are still linked to lexemes and generally have a less bleached meaning than derivational and inflectional morphemes, we hypothesize that they will be more subject to debonding than derivational and inflectional affixes. However, as shown by Van Goethem & De Smet (2012), other factors such as the strong degree of prosodical and morphological cohesion of Dutch compounds may counteract debonding in Dutch. (Provisional) results Our study will show that true derivational affixes in Dutch undergo debonding only very rarely. One exception, illustrated by Norde (2009: 213-220), is the case of the Dutch numeral suffix -tig (cf. twintig 'twenty', dertig 'thirty', veertig 'forty', etc.) which has recently developed new uses as a quantifier ('umpteen, dozens') (1) and as an intensifier ('much, very') (2). (1) Die kerel heeft al tig vriendinnen gehad. (Norde 2009: 213) 'That guy has already had dozens of girlfriends.' (2) beetje jammer, middelburg is toch tig leuker (Norde 2009: 217) 'bit of a shame, (the city of) Middelburg is much nicer after all' It will be argued that other cases such as the autonomous uses of ex 'ex; ex-wife, ex-husband', bi 'bisexual' and isme 'ism' should not be seen as the result of a gradual and context-sensitive debonding process, but as the result of abrupt and less context-sensitive lexicalization processes such as clipping and conversion. In some other cases, scope expansion of affixes without subsequent debonding may however be observed (e.g. ex-waarnemend burgemeester 'ex-acting mayor', anti-Michelle Martin betoging 'anti-Michelle Martin demonstration', kleine meisjesachtig 'lit. little girls-ish'). Compared to Dutch affixes, Dutch affixoids undergo debonding far more frequently. We will focus on Dutch intensifying prefixoids with a nominal origin that develop adjectival and/or adverbial uses through debonding. Examples are Dutch top 'lit. top; top-class, great', reuze 'lit. giant; great, very' and kei 'lit. boulder; very'. These three prefixoids can be used as autonomous adjectives, with attributive (3) and predicative (4) uses, and, in the case of reuze and kei, also as adverbs (5-6). (3) Verouderd hotel, maar top sfeertje hier! (NLCOW2012) 'Old-fashioned hotel, but top atmosphere here !' (4) Het aanbod in kwaliteitskamers is reuze. (NLCOW2012) 'The offer of quality rooms is gigantic.' (5) Eigenlijk valt het vasten reuze mee. (NLCOW2012) 'In fact, the fasting (diet) turns out much better than expected (lit. turns out giant).' (6) Je bent echt kei getalenteerd.(NLCOW2012) 'You are really talented (lit. boulder talented)' Our study will demonstrate that debonding of these prefixoids may be triggered and/or favored by (at least) four factors. First, the intensifying meaning of Dutch prefixoids plays an important role in their reanalysis as intensifying adjectives and/or adverbs. As Booij correctly observes, "[T]he meaning of intensification that is connected to these nouns is a type of meaning expressed prototypically by adjectives [and adverbs], and hence the categorial reinterpretation of these nouns as adjectives [and adverbs] in this context is a natural development" (Booij 2010: 61). Second, the form a many Dutch prefixoids, ending in -e (as nominal ending or linking morpheme in the compound pattern) (e.g. bere 'lit. bear; great, very', rete 'lit. ass; very', reuze 'lit. giant; great, very', klasse 'lit. class; classy', pokke 'smallpox; awful, very') may contribute to a reanalysis as a Dutch inflected adjective (e.g. een reuze stap 'a gigantic step', een erg klasse wijn 'a very classy wine'). Note that such a reanalysis is not possible with affixoids that primarily intensify adjectives (e.g. kei-), which may explain why these affixoids appear less subject to debonding. Third, it will be shown that debonding is facilitated when other processes, such as clipping and conversion, interact with it. More precisely, in the case of Dutch top 'top(-class), great', Van Goethem & Hüning (2013) have shown that its recent adjectival uses can be accounted for both by debonding of [top +N]N compounds (e.g. een top sfeertje 'a top atmosphere') and by N>A conversion in the predicative context (e.g. dat is echt (de) top 'that is really great (lit. (the) top)'). Whereas the first process suggests extension from the prefixoid use to the attributive and predicative uses (as suggested for Dutch by Booij 2010), the conversion pathway conversely suggests extension from the predicative uses to the attributive and prefixoid uses (as claimed for German by Pittner & Berman 2006). Besides, previous research has already indicated that clipping may also lead to adjectival uses of compound members (e.g. het feestje was reuze 'the party was great' < reuzeleuk 'lit. giant-fun') (Van Goethem & De Smet 2012). It is highly probable that debonding, on the one hand, and conversion or clipping, on the other hand, co-occur and mutually influence each other. Fourth, there is a strong tendency in Dutch (especially in informal writing) to write complex words (compounds and words containing affixoids) as two separate words, e.g. paarden lasagne instead of paardenlasagne 'horse lasagna', or dood saai instead of doodsaai 'dead boring'. These observations are completely in line with the recently developed idea of "multiple source constructions" (Van de Velde, De Smet and Ghesquière, forthc.), which claims that innovations in language change often result from (the interaction between) different sources and factors. The impact of the different factors (intensifying meaning, -e reanalysis, spelling) and processes (clipping, conversion) will be investigated into more detail by looking at the semantic, morphological and syntactic behavior (attributive, predicative and adverbial uses) of the autonomous uses of these prefixoids. Methodology The aforementioned factors and research questions will be applied to three case studies: top 'lit. top; top-class', reuze 'lit. giant; great, gigantic, very' and kei 'lit. boulder; very', all derived from nouns, occurring as intensifying prefixoids and manifesting 'debonded' uses as adjectives, and in the case of reuze and kei, also as adverbs. The data will be drawn from the COW (Corpora from the Web) internet corpus and will be submitted to an in-depth qualitative and quantitative analysis. References Booij, G. 2009. Compounding and Construction Morphology. In R. Lieber & P. Štekauer (Eds), The Oxford Handbook of Compounding, 201-216. Oxford: Oxford University Press. Booij, G. 2010. Construction Morphology. Oxford: Oxford University Press. COW2012 (Corpora from the web): http://hpsg.fu-berlin.de/cow/colibri/ Norde, M. 2009. Degrammaticalization. Oxford: Oxford University Press. Dressler, W.U., Kastovsky, D., Pfeiffer, O.E. and F. Rainer (Eds). 2005. Morphology and its Demarcations. Selected papers from the 11th Morphology meeting, Vienna, February 2004. Current Issues in Linguistic Theory 264. Amsterdam/Philadelphia: John Benjamins Publishing Company. Pittner, K. & J. Berman. 2006. video ist echt schrott aber single ist hammer-jugendsprachliche Nomen-Adjektiv-Konversion in der Prädikativposition. Deutsche Sprache 3. 233–250. Scalise, S. & I. Vogel (Eds). 2010. Cross-Disciplinary Issues in Compounding. Current Issues in Linguistic Theory 311. Amsterdam/Philadelphia: John Benjamins Publishing Company. Van de Velde, F., De Smet, H. & L. Ghesquière (Eds) (forthc.). Multiple source constructions in language change. Thematic issue in Studies in Language. Van Goethem, K & H. De Smet. 2012. How nouns turn into adjectives. The emergence of new adjectives in French, English and Dutch through debonding processes. Paper presented at the international conference on Adjectives in Germanic and Romance: Variation and Change, Amsterdam, March 2012. Van Goethem, K. & M. Hüning. 2013. Debonding of Dutch and German compounds. Paper presented at the international Germanic Sandwich Conference, Leuven, Jan. 2013.